Will determine the regulatory connectivity(s) in the PISK signaling network in a panel of luminal breast tumor cell lines using an unbiased systems approach and develop and test a robust mathematical model of the signaling to predict combinations for therapy. This project is motivated by the observation that the PISK network is selectively activated by mutation in luminal and HER2 positive tumors and represents the most frequent activating event in luminal breast cancers. PISK signaling regulates breast cancer proliferation, survival, cytokine production, protein translation, cell growth, bioenergetics and metastasis. Aberrations that activate PISK signaling can occur either extrinsically at the receptor level (HER2) or intrinsically within the network (RAS, PIKSCA, PIK3CR1, PDK1, AKT or PTEN). Furthermore, as demonstrated in this CCSB, the activation state ofthe PISK network determines outcomes to drugs targeting the HER2 network. Despite the evidence defining the PISK network as a high quality target, regulatory feedback loops as well as cross talk with other networks (MAPK) make understanding and targeting the PISK network a challenging task. One of the key goals will be develop robust predictive models able to determine rational combinatorial approaches to target aberrations in the PISK network and at the same time prevent deleterious effects of feedback loops and cross talk. Experimentally, this will be accomplished by determining PISK signaling network connectivity in 4 breast cell lines by targeted knockdown of all kinases followed by phospho-proteomic analysis. This connectivity network will be combined with the dynamic data from Project 1 to develop an ODE model that describes the AKT centric network response. The model can them be used to analyze patient data and determine combinations of drugs based on the specific signaling signature in the tumor tissue. We hypothesize that nested feedback loops in the PISK network and cross talk confer complex regulatory mechanisms that will need to be targeted. As a corollary, predictive computational models will have the power to identify rational drug combinations to effectively inhibit PISK network signaling and thus cell proliferation survival and metastasis. Models developed in this project will be informed by network development efforts in Project 1 and will contribute to development of models of response to MEK and HER2/3 inhibitors in Projects 2 and 3, respectively.